Low agricultural productivity is a persistent challenge in developing economies. In the first chapter of the dissertation, I study the concurrent but independently implemented expansion of rural roads and extension in Ethiopia to examine how access to markets and technologies affect agricultural productivity. Using geospatial data combined with large surveys and exploiting the staggered roll-out of the two programs, I show that there are strong complementarities between roads and extension. While ineffective in isolation, access to both a road and extension increases productivity. I find that roads and extension improve productivity by facilitating the take up of agricultural advice and modern inputs. Furthermore, households adjust crop choices and shift across occupations in response to their changing comparative advantages in access to markets and technologies. In the second chapter of the dissertation, co-authored with Samuel Bazzi and Martin Fiszbein, we study the long-run implications of the American frontier experience for culture and politics. We track the frontier throughout the 1790–1890 period and construct a novel, county-level measure of total frontier experience (TFE). Historically, frontier locations had distinctive demographics and greater individualism. Long after the closing of the frontier, counties with greater TFE exhibit more pervasive individualism and opposition to redistribution. We provide suggestive evidence on the roots of frontier culture: selective migration, an adaptive advantage of self-reliance, and perceived opportunities for upward mobility through effort. Overall, our findings shed new light on the frontiers persistent legacy of rugged individualism. In the third chapter of the dissertation, I use plant level census data to examine the effects of two policies designed to support prioritized sub-sectors and regions on the productivity of the Ethiopian manufacturing sector. The first policy, implemented during 1996-2002, was an activist industrial policy favoring import substitution while the second policy, active during 2003-2012, emphasized export promotion. I find that there is severe misallocation in Ethiopian manufacturing sector, but it has subsided over the studied period. The results suggest that the priority sector support policies have exacerbated the misallocation, and the within-sector variations of the policies largely account for the dispersion in revenue productivity.